Awaiting inside a loop runs the calls one after another: ten quotes of 200 ms each take two seconds. When the calls do not depend on each other, start them all first, then await them together with Task.WhenAll. The total wait is about as long as the slowest call. WhenAll returns the results in the order of the tasks you passed, not the order they finished.
If any task fails, await Task.WhenAll(tasks) throws the first exception only. The others are still there: check each task's IsCompletedSuccessfully, Result or Exception afterwards.
Task.WhenAny returns the first task to finish, which is how you race a call against Task.Delay for a timeout (or use task.WaitAsync(timeout), added in .NET 6).
Starting everything at once can overload the other side. A SemaphoreSlim with a count of 3 lets three calls in at a time: await gate.WaitAsync() before the call and gate.Release() in finally after it. Concurrency here means waiting together, not using more processor cores: these calls spend their time waiting for a network.
Start every call, await them together, and race a call against a delay.